• State TI Thinks Resolved
  • Locked Locked
  • Replies 17 replies
  • Answers 5 answers
  • Subscribers 62 subscribers
  • Views 1112 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

UCC25230: Secondary oscillation

shaowei cui
Intellectual 550 points
Part Number: UCC25230
Other Parts Discussed in Thread: PMP8878

Hi TI engineers,

I use UCC25230 as an auxiliary power source.Output 9V.Using LPD5030 as coupling inductor for second side power supply.But the oscillation is serious.The original edge is not oscillatory.

I do not konw why.

regars,

shaowei.

  • 0
    TI__Mastermind 26340 points
    Hello Shaowei,

    Please can you post schematic? Please can you detail your requirements?



    Thanks.


    David.
  • 0 in reply to David Baba
    Intellectual 550 points

    Hi David,

    This is the schematic.It uses UCC25230 as auxiliary power conversion chip converting from 36V-72V to 9V .The problem is that the T2 second side oscillation is serious as below channel 3.

     

    The schematic diagram is based on PMP8878 (TI reference design).How to eliminate oscillations？

    Unlike the reference design, the T2 model I used was LPD5030V-224MEB whose DCR is 3.3 Ohms, while the reference design used LPD5030V-224MRB whose DCR is 5.25 Ohms.

    Is the oscillation caused by a smaller DC equivalent resistance?

    Regars,

    Shao wei.

  • 0 in reply to shaowei cui
    TI__Mastermind 26340 points
    Hello Shao,

    Please can you measure the Switch node on the primary Side which is Pin 1 of the device. Also I see you have an input filter that will probably need to be damped. Please see link below on how to implement parallel damping. In short add a high capacitance electrolytic with high ESR in parallel with the existing input caps, place by right side of L2.


    www.ti.com/.../snva538.pdf




    David.
  • 0 in reply to David Baba
    Intellectual 550 points

    Hi David,

        

    switch node at 15V input.

    switch node at 30V input.

    switch node at 40V input.

    The electrolytic capacitor can not be connected in parallel, because this is the 1/4 brick module power source without enough space.

    This input filter is based on PMP8878 reference design whose input voltage range is 40-75v,can I use these parameters since there was no oscillation from their test report ?

    regads,

    shaowei.

  • 0 in reply to shaowei cui
    TI__Mastermind 26340 points

    Hello Shaowei,

    It looks to me that the waveform's you sent on your original post is not an oscillation due to instability as shown by your switch node waveform's, but its noise. The source of the noise is the ringing on the switching edge seen at the anode side of the secondary diode. This high frequency ringing end up on the output due to the output capacitor being unable to filter due to its self-resonance.

    Where you place your ground of your scope probe is also of great importance. I suggest placing a Spring clip on the barrel of the scope probe to prevent noise pick up and kelvin connect directly across the output capacitor.

    Also layout is of great importance, the Diode and output capacitor on the secondary needs to be positioned as close as possible to the secondary side of the transformer and form a tight loop from Diode cathode to Cout positive and Cout ground back to the dot of the secondary of the transformer.

    If the layout is as described above you maybe able to further filter out by placing low value ceramics to decouple the noise on the output, suggest 100nF, 10nF and a 1nF in parallel. If it is layout related, you will not be able to and placing decoupling caps will prove to be ineffective in resolving the issue.

    If this is the case, you may have to re-position as described above and retest?

    hope this helps?

    David.

  • 0 in reply to David Baba
    Intellectual 550 points

    Hi David,

    Paralleling low capacitance values do not work.I will rearrange the layout.

    There is another question . Why does output increase with the increase of input?

    Is it because of the oscillating spike that charges the voltage up?

    Input at 15V.

    Input at 60V.

    Input at 72V.

  • 0 in reply to David Baba
    Intellectual 550 points

    Hi Peter,

    This is the PCB layout.Is the circuit made of transformer, diode and output capacitor too large?

    Regars,

    shaowei

  • 0 in reply to shaowei cui
    TI__Mastermind 26340 points

    Hello Shaowei,

    What is your input voltage range?  How much does you output vary with a given change on Vin?  With a FlyBuck, the output will always move a little because the regulator is controlling the primary voltage and not the secondary voltage.

    Thanks.

    David.

  • 0 in reply to shaowei cui
    TI__Mastermind 26340 points
    Hello Shaowei,

    The layout could be improved a little on the secondary as I described above. Also the primary side components are not well positioned. You should have the FB resistors and compensation components close to the device with the Gnd going back to the devices Gnd pin with a short path. Also Boot cap and VCC caps needs to be close to device pins.

    Hope this helps?

    David.
  • 0 in reply to David Baba
    Intellectual 550 points
    Hi David,
    The input voltage range is 36-72V.The secondary output vary from 11V to 18V,but primary is still 9V.I don't know why.
    You mean the noise is caused by primary layout?

    Regards,
    shaowei.
  • 0 in reply to David Baba
    Intellectual 550 points

    Hi David,

    How about this time?Is it OK?

    Regars,

    shaowei

  • 0 in reply to shaowei cui
    Intellectual 550 points

    Hi David,

    How about this time,is it OK?

    Regars,

    shaowei

  • 0 in reply to shaowei cui
    TI__Mastermind 26340 points

    Hello Shaowei,

    Yes, layout is critical on both primary and secondary.  I am assuming there is no load on the secondary when you see the issue with output voltage increasing with input voltage? If so, you need to pre-load the secondary in application where the secondary is unloaded.  This is because you have ringing on the secondary that peak charges the output cap on the secondary and this will creep up.  As Vin is increased the voltage ring on the secondary will get worse further peak charging the output cap to a greater voltage?

    Either pre-load secondary with a resistor that draws a couple of mAs of current or place a zener clamp?

    Hope this helps?

    David.

  • 0 in reply to David Baba
    TI__Mastermind 26340 points

    See attached suggested layout.

    UCC25230 Floor plan.pdf

    Hope this helps?

    David.

  • 0 in reply to David Baba
    Intellectual 550 points
    Hi David,
    Thank you very much!
    Regars,
    shaowei
  • 0 in reply to David Baba
    Intellectual 550 points
    Hi David,
    After I pre-loading the secondary and placing a Spring clip on the barrel of the scope probe to prevent noise pick up and kelvin connect directly across the output capacitor,The secondary output is kept at 9V and no longer oscillates.That is to say, the previous problem was caused by measurement.
    What is noise?Will the noise affect the load, that is, the controller and the GaN driver?
    Because the maximum GaN drive voltage can not exceed 6V, I want to know whether the secondary output is oscillating in the end to ensure reliable driving of GaN?
    Regars,
    shaowei
  • 0 in reply to shaowei cui
    TI__Mastermind 26340 points

    Hello Shaowei,

    A GAN driver has an Internal Bootstrap Supply Voltage Clamping the supply to ensure that the voltage does not exceed limits set by the GAN devices.

    You should not have an issue here.

    Hope this helps?

    David